pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis

TL;DR

pi-GAN introduces a novel 3D-aware image synthesis model using periodic activations and volumetric rendering, achieving state-of-the-art multi-view consistent results with high detail.

Contribution

The paper presents pi-GAN, a new generative model that improves 3D-aware image synthesis by using periodic activations and volumetric rendering for better view consistency and image quality.

Findings

Achieves state-of-the-art results on multiple datasets.

Produces view-consistent 3D representations with fine detail.

Outperforms existing methods in 3D-aware image synthesis.

Abstract

We have witnessed rapid progress on 3D-aware image synthesis, leveraging recent advances in generative visual models and neural rendering. Existing approaches however fall short in two ways: first, they may lack an underlying 3D representation or rely on view-inconsistent rendering, hence synthesizing images that are not multi-view consistent; second, they often depend upon representation network architectures that are not expressive enough, and their results thus lack in image quality. We propose a novel generative model, named Periodic Implicit Generative Adversarial Networks ($\pi$-GAN or pi-GAN), for high-quality 3D-aware image synthesis. $\pi$-GAN leverages neural representations with periodic activation functions and volumetric rendering to represent scenes as view-consistent 3D representations with fine detail. The proposed approach obtains state-of-the-art results for 3D-aware image synthesis with multiple real and synthetic datasets.